Preliminary results suggest you both may be correct.
As I indicated in an earlier post, mine is an attempt to apply both Skinner's and Pavlov's findings regarding classical conditioning to a relatively simple personal scenario, i.e., invasive squirrels. Thusfar, I have noted a marked reluctance, on the part of a regular cohort of squirrels living in my back yard trees, to attempt to mount the cardinal feeder and consume the sunflower seeds within, after a period of approximately one month wherein the offenders were subjected to various body shots with the Nerf® gun from an average engagement distance of ~10 feet.
Initially, the squirrels were brazen, returning repeatedly to feed within minutes of successful engagement with the gun (head and body contact with projectiles.) Currently, I have noticed behavior on the part of the rodents indicating a heightened sensitivity to the visual and audatory cues which have preceded the gunfire, e.g., flinching and stutter steps when I approach the window, and the sound of my wood floors creaking as I walk to grab the gun, etc. Often, this is enough to send them packing. However, if I leave the house for even half an hour, I inevitably arrive home to see one of two of the scoundrels on the feeder.
In addition, I have noted on a few occasions that squirrels will approach the feeder, pause momentarily as if weighing this risk/reward scenario, then "decide" to bail--before I have made any sort on motion to engage them.
I agree that the drive for nutrition is a strong one; we'll see how this all plays out once I'm no longer here each day to monitor them because of school.